Intracardiac echocardiography: computerized detection of left ventricular borders

Comparative studies of deep learning segmentation models for left ventricle segmentation

One of the primary factors contributing to death across all age groups is cardiovascular disease. In the analysis of heart function, analyzing the left ventricle (LV) from 2D echocardiographic images is a common medical procedure for heart patients. Consistent and accurate segmentation of the LV exerts significant impact on the understanding of the normal anatomy of the heart, as well as the ability to distinguish the aberrant or diseased structure of the heart. Therefore, LV segmentation is an important and critical task in medical practice, and automated LV segmentation is a pressing need. The deep learning models have been utilized in research for automatic LV segmentation. In this work, three cutting-edge convolutional neural network architectures (SegNet, Fully Convolutional Network, and Mask R-CNN) are designed and implemented to segment the LV. In addition, an echocardiography image dataset is generated, and the amount of training data is gradually increased to measure segmentation performance using evaluation metrics. The pixel's accuracy, precision, recall, specificity, Jaccard index, and dice similarity coefficients are applied to evaluate the three models. The Mask R-CNN model outperformed the other two models in these evaluation metrics. As a result, the Mask R-CNN model is used in this study to examine the effect of training data. For 4,000 images, the network achieved 92.21% DSC value, 85.55% Jaccard index, 98.76% mean accuracy, 96.81% recall, 93.15% precision, and 96.58% specificity value. Relatively, the Mask R-CNN outperformed other architectures, and the performance achieves stability when the model is trained using more than 4,000 training images.

Ultrasonic sensor concept to fit a ventricular assist device cannula evaluated using geometrically accurate heart phantoms

Future left ventricular assist devices (LVADs) are expected to respond to the physiologic need of patients; however, they still lack reliable pressure or volume sensors for feedback control. In the clinic, echocardiography systems are routinely used to measure left ventricular (LV) volume. Until now, echocardiography in this form was never integrated in LVADs due to its computational complexity. The aim of this study was to demonstrate the applicability of a simplified ultrasonic sensor to fit an LVAD cannula and to show the achievable accuracy in vitro. Our approach requires only two ultrasonic transducers because we estimated the LV volume with the LV end-diastolic diameter commonly used in clinical assessments. In order to optimize the accuracy, we assessed the optimal design parameters considering over 50 orientations of the two ultrasonic transducers. A test bench was equipped with five talcum-infused silicone heart phantoms, in which the intra-ventricular surface replicated papillary muscles and trabeculae carnae. The end-diastolic LV filling volumes of the five heart phantoms ranged from 180 to 480 mL. This reference volume was altered by ±40 mL with a syringe pump. Based on the calibrated measurements acquired by the two ultrasonic transducers, the LV volume was estimated well. However, the accuracies obtained are strongly dependent on the choice of the design parameters. Orientations toward the septum perform better, as they interfere less with the papillary muscles. The optimized design is valid for all hearts. Considering this, the Bland-Altman analysis reports the LV volume accuracy as a bias of ±10% and limits of agreement of 0%-40% in all but the smallest heart. The simplicity of traditional echocardiography systems was reduced by two orders of magnitude in technical complexity, while achieving a comparable accuracy to 2D echocardiography requiring a calibration of absolute volume only. Hence, our approach exploits the established benefits of echocardiography and makes them applicable as an LV volume sensor for LVADs.

The use of intracardiac echocardiography during percutaneous pulmonary valve replacement

High-quality live imaging assessment of cardiac valves and cardiac anatomy is crucial for the success of catheter-based procedures. We present our experience using Intracardiac echocardiography (ICE) during transcatheter Percutaneous Pulmonary Valve replacement (tPVR).This is a retrospective study that included 35 patients who underwent tPVR between April 2008 and June 2012. Thirty-one of these patients had the procedure performed under continuous ICE guidance. Pre-procedure transthoracic echocardiography (TTE) was obtained in all patients. ICE was performed at baseline, during the procedure, and at the conclusion of the procedure. Comparisons between the pre-procedure TTE and baseline ICE data and between post-procedure ICE data and the following day TTE were performed. Total of 35 patients had tPVR during the above-mentioned time period. Twenty-one patients received the Edwards Sapien valve and 14 patients had the Melody valve. Thirty-one patients had the procedure performed under continuous ICE guidance. The mean Pre-TTE peak gradient (PG) and Pre-ICE-PG were 45.5 ± 20 vs 33 ± 13 mmHg (p < 0.001) and the mean Pre-TTE mean gradient (MG) and Pre-ICE-MG were 27.7 ± 13 vs 21 ± 18 mmHg (p < 0.001). The mean Post-TTE- PG and Post-ICE-PG were 24.3 ± 11 vs 15.3 ± 7 mmHg (p < 0.001) and the mean of the Post-TTE-MG and Post-ICE-MG were 14.2 ± 7 vs 8.4 ± 4 mmHg (p < 0.001). There was a good correlation between peak ICE and TTE gradient at baseline and after valve placement. For the degree of pulmonary regurgitation, there was no significant difference between TTE and ICE. ICE is an important modality to guide tPVR in patients with dysfunctional homograft valve between the right ventricle and pulmonary artery and should be used to assess valve function before, during and immediately after the procedure.

Impact of inactivity and exercise on the vasculature in humans

The effects of inactivity and exercise training on established and novel cardiovascular risk factors are relatively modest and do not account for the impact of inactivity and exercise on vascular risk. We examine evidence that inactivity and exercise have direct effects on both vasculature function and structure in humans. Physical deconditioning is associated with enhanced vasoconstrictor tone and has profound and rapid effects on arterial remodelling in both large and smaller arteries. Evidence for an effect of deconditioning on vasodilator function is less consistent. Studies of the impact of exercise training suggest that both functional and structural remodelling adaptations occur and that the magnitude and time-course of these changes depends upon training duration and intensity and the vessel beds involved. Inactivity and exercise have direct "vascular deconditioning and conditioning" effects which likely modify cardiovascular risk.

Echocardiography-guided interventions

A major advantage of echocardiography over other advanced imaging modalities (magnetic resonance imaging, computed tomographic angiography) is that echocardiography is mobile and real time. Echocardiograms can be recorded at the bedside, in the cardiac catheterization laboratory, in the cardiovascular intensive care unit, in the emergency room-indeed, any place that can accommodate a wheeled cart. This tremendous advantage allows for the performance of imaging immediately before, during, and after various procedures involving interventions. The purpose of this report is to review the use of echocardiography to guide interventions. We provide information on the selection of patients for interventions, monitoring during the performance of interventions, and assessing the effects of interventions after their completion. In this document, we address the use of echocardiography in commonly performed procedures: transatrial septal catheterization, pericardiocentesis, myocardial biopsy, percutaneous transvenous balloon valvuloplasty, catheter closure of atrial septal defects (ASDs) and patent foramen ovale (PFO), alcohol septal ablation for hypertrophic cardiomyopathy, and cardiac electrophysiology. A concluding section addresses interventions that are presently investigational but are likely to enter the realm of practice in the very near future: complex mitral valve repairs, left atrial appendage (LAA) occlusion devices, 3-dimensional (3D) echocardiographic guidance, and percutaneous aortic valve replacement. The use of echocardiography to select and guide cardiac resynchronization therapy has recently been addressed in a separate document published by the American Society of Echocardiography and is not further discussed in this document. The use of imaging techniques to guide even well-established procedures enhances the efficiency and safety of these procedures.

Artificial neural network: border detection in echocardiography

Being non-invasive and low cost, the echocardiography has become a diagnostic technique largely applied for the determination of the left ventricle systolic and diastolic volumes, which are used indirectly to calculate the left ventricle ejection volume, the cardiac cavities muscular contraction, the regional ejection fraction, the myocardial thickness, and the ventricular mass, etc. However, the image is very noisy, which renders the delineation of the borders of the left ventricle very difficult. While there are many techniques image segmentation, this work chooses the artificial neural network (ANN) since it is not very sensitive to noise. In order to reduce the processing time, the operator selects the region of interest where the neural network will identify the borders. Neighborhood and gradient search techniques are then employed to link the points and the left ventricle contour is traced. The present method has been efficient in detecting the left ventricle borders echocardiography images compared to those whose borders were delineated by the specialists. For good results, it is important to choose properly the areas to be analyzed and the central points of these areas.

Segmentation in Echocardiographic Sequences using Shape-based Snake Model Combined with Generalized Hough Transformation

A novel method for segmentation of cardiac structures in temporal echocardiographic sequences based on the snake model is presented. The method is motivated by the observation that the structures of neighboring frames have consistent locations and shapes that aid in segmentation. To cooperate with the constraining information provided by the neighboring frames, we combine the template matching with the conventional snake model. It means that the model not only is driven by conventional internal and external forces, but also combines an additional constraint, the matching degree to measure the similarity between the neighboring prior shape and the derived contour. Furthermore, in order to auto or semi-automatically segment the sequent images without manually drawing the initial contours in each image, generalized Hough transformation (GHT) is used to roughly estimate the initial contour by transforming the neighboring prior shape. The method is particularly useful in case of the large frame-to-frame displacement of structure such as mitral valve. As a result, the active contour can easily detect the desirable boundaries in ultrasound images and has a high penetrability through the interference of various undesirables, such as the speckle, the tissue-related textures and the artifacts.

Use of novel nonfluoroscopic three-dimensional electroanatomic mapping system to monitor and analyze heart surgery in animal models

BACKGROUND

The new method of three-dimensional (3D) electroanatomic mapping was presented as an important tool for cardiac imaging and intervention. We present herein the first use of this technology for the monitoring, analysis, and development of cardiac surgery at the preclinical stage.

METHODS

The method is based on utilizing a locatable catheter connected to an endocardial mapping and navigating system, to accurately establish the location and orientation of the tip of the mapping catheter and simultaneously record its local electrogram. The 3D geometry of the beating cardiac chamber is reconstructed in real time. The system was tested on six goats that underwent dynamic cardiomyoplasty. Two maps of each animal were performed: preoperative and postoperative during the stimulation protocol of the skeletal muscle.

RESULTS

The electroanatomic mapping system provided detailed maps of the left ventricle during the stimulation protocol, which demonstrated a striking geometric difference between the assisted and the unassisted beats. These geometric changes are best described by referring to left ventricular long-axis movements (22.3 +/- 3.8 degrees vs 3.4 +/- 1.6 degrees, p < 0.001), center-of-mass movements (10.4 +/- 3.0 mm vs 3.9 +/- 1.6 mm, p < 0.005), and the changes in upward movement viewed along the base (7.9 +/- 1.9 mm vs 3.6 +/- 1.7 mm, p < 0.01), middle (13.8 +/- 4.0 mm vs 7.3 +/- 1.8 mm, p < 0.005), and the apex of the heart (28.1 +/- 4.5 vs 5.3 +/- 2.3 mm, p < 0.001) [mean +/- SD].

CONCLUSIONS

The 3D electroanatomic mapping system allows detailed reconstruction of the left ventricular geometry and a clear view of the difference between the assisted and the unassisted beats. This novel monitoring system may serve as an important tool for the analysis and development of new techniques in cardiac surgery.

Use of Intracardiac Echocardiography to Guide Catheter Closure of Atrial Communications

Intracardiac echocardiography (ICE) is slowly replacing transesophageal echocardiography as the preferred imaging tool to guide device closure of atrial septal defects and patent foramen ovale. This article is a brief review of the literature related to ICE, the technical aspects ICE imaging, techniques for obtaining the standard views, and the future directions of this methodology.

Improved analysis of brachial artery ultrasound using a novel edge-detection software system

Brachial artery ultrasound is commonly employed for noninvasive assessment of endothelial function. However, analysis is observer dependent and susceptible to errors. We describe studies on a computerized edge-detection and wall-tracking software program to allow more accurate and reproducible measurement. In study 1, three purpose-built Perspex phantom arteries, 3.00, 4.00, and 6.00 mm in diameter, were measured with the software. There was a mean bias of 11 microm (P < 0.001 at each level) between known and measured values; the mean resolving power of the software was estimated as 8.3 microm. In study 2, the mean intraobserver coefficient of variation of repeated measures of flow-mediated dilation (FMD) using the software (6.7%) was significantly lower than that for traditional manual measurements using the intima-lumen interfaces (24.8%, P < 0.05) and intima-media interfaces (32.5%, P < 0.05). In study 3, 24 healthy volunteers underwent repeat testing twice within 1 wk; the coefficients of variation for between-visit reproducibility of FMD and response to glyceryl trinitrate using the software were 14.7 and 17.6%, respectively. Assuming 80% power and an alpha of 0.05, eight subjects with matched controls would be required, in a parallel designed study, to detect an absolute 2.5% change in FMD. In summary, we have developed a semiautomated computerized vascular ultrasound analysis system that will improve the power of clinical intervention studies to detect small changes in arterial diameter.